Functional Recovery Ten Years after Pediatric Traumatic Brain Injury: Outcomes and Predictors

Abstract

Functional impairments (adaptive, behavioral, educational) are common after preschool traumatic brain injury (TBI). In comparison with cognitive outcome, functional outcomes have received limited attention, with little evidence to determine whether these difficulties persist in the long term. The aim of this study was to examine functional outcomes at 10 years post-injury and identify predictors of outcome. The study compared children with a diagnosis of TBI (n=40) to a healthy age-, gender-, and socioeconomic status (SES)-matched control group (n=19) at 10 years post-injury. Outcomes and predictors of functional skills were investigated. Poorer adaptive skills were evident for those with more severe injury. Behavioral difficulties were present regardless of injury severity. Post-injury, arithmetic skills were the most compromised in the longer term. Pre-injury status, interventions accessed, and acute intellectual function were significant predictors of outcome. These results highlight the importance of monitoring functional skills in the long term, especially for those children presenting with risk factors.

Introduction

R esearch investigating long-term survivors of traumatic brain injury (TBI) indicates that it is functional impairments, including adaptive (e.g., practical, conceptual) and behavioral (e.g., internalizing and externalizing behaviors, behavioral symptoms) disturbances and poor educational progress (e.g., literacy), that can cause distress and lead to greatest difficulties in adjustment post injury.^1
–3 In contrast to physical and cognitive symptoms, which tend to reduce or stabilize over time,⁴ these functional difficulties often persist, and even increase over time.^5

–8 In the longer term, the burden of these problems increases the risk of psychiatric disorders (e.g., depression, anxiety),^9,10 especially after more severe insult and where family dysfunction is evident.^11
–13

In contrast to the substantial literature addressing cognitive deficits, the functional consequences of TBI have received less attention.^13

–18 Recent literature is attempting to fill this gap, with functional impairments (adaptive, behavioral and educational) investigated in the shorter term following childhood TBI in school-aged populations.^{5,11,13,19,20} To date, however, little research has included a population younger than early primary school children, in whom injury occurred during a period of rapid brain development and cognitive skills maturation, and who, therefore, may be most vulnerable to the disruptions caused by a brain injury. Most studies addressing this younger age group have been cross-sectional, but results have suggested functional deficit even after mild injury.^21

–24

Researchers have previously reported on acute and subacute functional outcomes following TBI during early childhood.^1,25,26 Anderson et al.^25,26 examined three groups of children sustaining injuries of different severity and compared them with a healthy control group. Using a prospective, longitudinal design, adaptive abilities, behavior, educational progress, and everyday memory skills were investigated acutely post-injury and again at 12 and 30 months post-injury. Results suggested a strong association between injury severity and outcomes across all domains, with 30 month functional outcome predicted by injury severity, family factors, and pre-injury levels of child function. The sample was further examined at 5 years,¹ and a strong association was again found between injury severity and adaptive, behavioral, and educational areas, with children who had sustained greater injury being most at risk. Furthermore, although educational performance was best predicted by injury severity, 5 year outcomes in adaptive and behavioral areas were best predicted by pre-injury levels of child function, with no significant difference in SES in the pre-injury stage to confound the predictive power of pre-injury level of function.

More recently Wells et al.²⁷ investigated social and functional outcomes using measurement tools in the areas of occupational therapy, physiotherapy, and psychology, in a group of children injured between birth and 10 years of age. These authors supported the idea that a combination of factors such as age at injury, clinical rating of injury severity, and environmental factors (e.g., nurturing environment) provided the best estimate of long-term functional outcome. Gerring et al.²⁸ also focused on behavioral outcome post-TBI, with an interest in pre-injury prevalence and post-injury incidence of Diagnostic and Statistical Manual of Mental Disorders, 3rd edition, revised (DSM-III-R) oppositional defiant disorder (ODD) and conduct disorder (CD). The authors concluded that pre-injury CD is a significant risk factor for behavioral maladjustment, and that new-onset CD and the associated behaviors are present at 1 year post-injury.

Chapman et al.⁵ and Yeates et al.⁸ investigated functional outcomes in the initial 18 months post-TBI in children injured between the ages of 3 and 7 years, and compared them with a group of children with orthopedic injuries. Results suggested that, with regard to behavioral outcomes over the 18 month period, the group that sustained severe injury developed significantly more externalizing problems than did the control group. As in school- age groups, family dysfunction, permissive parenting, and low SES were found to be significant predictors of outcome. Furthermore, although the family environment does moderate outcome, this influence may lessen in time for those children who sustained a severe injury.

On balance, the literature suggests that those with more severe injury present with functional deficits in the short and longer term post-injury, whereas for those sustaining a mild injury, symptoms resolve by 3–6 months post-injury.²⁹ The literature presented previously highlights commonalities among articles investigating functional outcomes post-childhood TBI, with an indication that these deficits are evident, especially following severe TBI. Furthermore, pre-injury, age at injury, injury, and environmental factors all have a role in the prediction of both shorter and longer-term outcome.^{7,12,15,18,22,25,26,30

–35}

The present study aimed to: 1) address the impact of TBI during early childhood on functional skills (adaptive ability, behavior, and educational performance) at 10 years post-injury; and 2) to identify predictors of functional recovery at 10 years post-TBI. The unique features of this study, in comparison with previous literature, are the inclusion of a TBI sample of children younger than those in early primary school and a matched control group, and a prospective longitudinal design following children to 10 years post-injury. To date, no other study has serially assessed this age group in a systematic manner up to 10 years post-insult. It was hypothesized, first, that more severe TBI would be associated with poorer and more persistent functional deficits in all domains. In addition, it was expected that injury severity, acute intellectual function, premorbid child ability levels (adaptive function, behavior), pre-injury family factors, and social/environmental factors (SES, intervention services used over the 10 years post-insult), would influence recovery and outcome in the 10 year post-injury period.

Methods

Participants

The present study describes data collected as part of a prospective longitudinal study of young children with TBI.^36,37 During the initial recruitment period, 109 children were admitted to the Royal Children's Hospital (RCH), Melbourne, Australia, with a diagnosis of TBI. Of these, the original sample for the longitudinal study comprised 96 children with TBI. These children had a diagnosis of closed head injury and were recruited from consecutive admissions to the neurosurgical ward at RCH, between June 1993 and June 1997, immediately following their injury. Inclusion criteria were: 1) age at injury 1.0–8.00 years; 2) documented evidence of TBI on presentation to the Emergency and Intensive Care Units of the RCH, including evidence of a blow to the head and a period of altered consciousness; and 3) medical records sufficiently detailed to determine injury severity - Glasgow Coma Score (GCS)³⁸ period of unconsciousness/coma (GCS≤8), imaging results where available; and 4) English speaking. Exclusion criteria were: 1) TBI as a result of child abuse; 2) penetrating head injury; 3) documented history of previous TBI; and 4) evidence of pre-existing learning, behavioral, physical, neurological, psychiatric, or developmental disorder.

Forty children with TBI, (42%) of the original sample, participated in data collection for the 10 year follow-up study. Of note was that high attrition rates are common in longitudinal research,^31,39 particularly in the field of TBI, and in young samples, where families are more likely to be mobile and difficult to locate with time. By using a prospective cohort, rather than a cross-sectional design, we were able to examine the representative nature of the 10 year sample in contrast to the initial study cohort on key demographic variables that might impact outcome (e.g., age, SES).

Nineteen children (54% of the original control sample) comprised the uninjured control group. The remainder of the control sample was unable to be located. These healthy children had been identified 10 years earlier via preschools and childcare centers, during the initial stage of the study, to match the TBI group as closely as possible with respect to age, gender, parent occupation, and pre-injury characteristics. Therefore, the control group was matched to the TBI group with respect to sociodemographic factors. Often, the cause of injury in the age group included in this study is closely related to proximal and distal social factors such as SES, family environment, parental supervision, and parent education.¹ Whereas orthopedic controls may be relevant for older children to control for behavioral traits, for example, risk taking behavior, in younger children these injuries are more likely to be the result of parent factors (e.g., drunk driving, poor supervision), with TBI more commonly linked to sociodemographic factors.^25,36 Control children have been reviewed, along with children with TBI, at all time points through the study. Inclusion criteria 1) and 4), and exclusion criteria 3) and 4), described previously, also applied to controls.

Children with TBI were divided into severity groups based on several measures, as no single measure has been found to be reliable in this age group. Acute CT/MRI scans were reported by a pediatric neuroradiologist and neurosurgeon, with classification of pathology (frontal, extrafrontal, diffuse) determined on the basis of radiological reports. Severity groups were derived as follows: 1) mild TBI (n=7): GCS on admission 13–15,³⁸ indicating some alteration of consciousness level (e.g., drowsiness, disorientation), with no evidence of mass lesion on CT)/MRI), and no other diagnosed neurologic deficits (e.g., visual difficulties, hearing loss, motor problems, seizures); 2) moderate TBI (n=20): GCS on admission 9–12, indicating significantly altered consciousness, with reduced responsiveness; and/or mass lesion or other evidence of specific injury on CT/MRI, and/or neurological impairment; and 3) severe TBI (n=13): GCS on admission≤8, representing coma, and mass lesion or other evidence of specific injury on CT/MRI, and/or neurological impairment. Implementation of this categorization procedure for severity successfully classified most children with TBI; however, where categorization was not clear, further information from the child's medical file (e.g., presence of neurological signs) was taken into account.

Materials

Although the sample was evaluated at six time points: acute (0–3 months post-injury), 6, 12, and 30 months, and 5 and 10 years post-injury, for the current study, the focus of the current article is the 10 year time point.

Injury and demographic variables

Data were collected on each child's medical and developmental history, parent occupation, and family constellation, where an intact family was defined as a two parent family, married and living in the same home. During inpatient stay, medical records were reviewed daily and GCS, length of coma, neurological abnormalities, and surgical interventions were recorded. GCS on admission was recorded by the admitting medical officer. SES was coded using Daniel's Scale of Occupational Prestige,⁴⁰ which rates parent occupation on a seven point scale, where a high score represents low SES. Data were also collected for the children with TBI regarding the kinds of services, such as treatment/intervention and medication, that they received throughout the 10 years, and which are provided in Table 4. No child was on medication at 10 years post-injury.

Parent questionnaires

Families agreeing to participate completed the following questionnaires at acute (0–3 months post-injury) and 10 years post-injury time-points.

Adaptive functioning

1) The Vineland Adaptive Behavior Scale (VABS)⁴¹ was administered on recruitment to the study (0–3 months post-injury), and parents were asked to rate their child's pre-injury behavior. Communication, Daily Living Skills, Socialization, and a Total Adaptive Behavior scores (VABS:TOT) were derived (mean=100, SD=15), and used in analyses during the acute stage post-injury. 2) At 10 years post-injury, adaptive behavior was assessed using the Adaptive Behavior Assessment System – Second Edition (ABAS-II)⁴² parent version, Conceptual, Practical and Social, and a Total Composite score (mean=100, SD=15) were obtained and used as the outcome measures for the adaptive function domain at 10 years post-injury. Impairment rates at 10 years post-injury (i.e., scores>1 SD below the mean) were also calculated and used in analyses.

Behavioral functioning

The revised format version of the Personality Inventory for Children (PIC)⁴³ was employed, which included 131 items for which parents respond either true or false. Parents completed this questionnaire on recruitment, based on their child's pre-injury level of functioning. Four factors (mean=50, SD=10) are derived from the scale: Undisciplined/Poor Self-control; Social Incompetence; Internalization/Somatic Symptoms; and Cognitive Development. Scores of ≥70 represent behavioral difficulties of clinical significance. At 10 years, the Behavior Assessment System for Children (BASC),⁴⁴ parent version was administered. The BASC provides Externalizing, Internalizing and Behavioral Symptoms Indices (mean=50, SD=10). A score of ≥60 is considered clinically significant. Impairment rates (i.e., scores>60) were also investigated at 10 years post-injury.

Family functioning

Family Functioning Questionnaire (FFQ),⁴⁵ was completed during the acute stage post-injury as a pre-injury measure and again at 10 years post-injury. Each item was rated on a six point scale, ranging from 1=totally agree to 6=totally disagree. Three factors are derived from the measure: Conflict (scored out of 60 points); Intimacy (72 points), and Parenting Style (30 points). For each factor, a higher score reflects more of that characteristic reported by families. The Intimacy factor was chosen for inclusion in statistical analyses, because of its high correlation, in our sample, with both other factors, and is comparable to a family cohesion factor.

The Family Burden of Injury Scale (FBIS)³⁴ was completed at 10 years post-injury only. This measure is a 27 item questionnaire that provides a Total Burden Score, and five subscales are derived from the scale: Child, Spouse, Others, Siblings, and Family routines/planning. Total Burden Score was used in analyses.

Child evaluations

Intellectual abilities

The Bayley Scales of Infant Development (BSID)⁴⁶ or the Wechsler Preschool and Primary Intelligence Scale-Revised (WPPSI-R)⁴⁷ or the Wechsler Intelligence Scale for Children - III (WISC-III)⁴⁸ was administered, depending upon the age of child at the acute stage post-injury (BSID<2.5 years, WPPSI-R<6.5 years, WISC-III≥6.5 years) during the acute stage post-injury (0–3 month, depending upon the child's recovery). The WISC-III⁴⁸ or the Wechsler Adult Intelligence Scale - Third Edition (WAIS)⁴⁹ was administered, depending upon the age of the child at 10 years post-insult. Full Scale Intelligence Quotient (FSIQ; mean=100, SD=15) was used in analyses at acute and 10 year data points.

Educational skills

The Wide Range Achievement Test - 3. (WRAT-3)⁵⁰ was administered at 10 years post-TBI and assessed reading, spelling, and arithmetic skills, (mean=100, SD=15). Impairment rates were also calculated at 10 years post-injury, where a score of >1 SD below the mean indicated impairment.

Procedure

The study was approved by the Human Research Ethics Committee of the RCH, Melbourne, Australia. Children were enrolled in the study during their initial hospital admission. Families were given a detailed description of the study and asked to provide written informed consent, in keeping with hospital ethics procedures. Once they had agreed to participate, parents were requested to complete the demographic questionnaires based on the child's pre-injury abilities, and then again at each assessment point, based on current function.

Children were evaluated at six data points: acute (0–3 months post-injury), 6, 12, and 30 months, and 5 and 10 years post-injury. The 10 year time point is the focus of this article, which is when each assessment occurred over two 1 h sessions. A qualified child psychologist conducted assessments on an individual basis, with intelligence quotient (IQ) measures administered in the first session and educational measures administered in the second session. The order of test presentation was fixed within each session.

Statistical analysis

Initially, the four groups (mild, moderate, severe, controls) were compared on demographic, pre-injury and psychosocial measures to identify any differences across groups that might influence post-injury performance. Analysis of variance (ANOVA) with planned contrasts was then conducted, covarying for SES at the acute stage post-injury and again at 10 years, when raw scores were employed (only in analyses where these were significant covariates), comparing those children with more severe injuries to those with lesser or no brain injury, on functional measures at 10 years post-injury. More specifically, for Contrast 1, the group with severe TBI was compared with the group with mild TBI and the control group combined, and for Contrast 2, the group with moderate TBI was compared with the group with mild TBI and the control group combined, hypothesising that the group with more severe insult would perform poorest. Effect sizes (ES) using Cohen's d ⁵¹ were calculated and, according to Tabachnick and Fidell⁵¹ were defined as small, d=0.2, medium, d=0.5, and large, d=0.8, and an ES of 0.67 was considered clinically significant.

Chi-square analysis was used to investigate frequency of functional impairment. A cutoff of one SD from the expected mean was employed to identify those children who were presenting with some degree of difficulty, consistent with that identified as having “functional significance.”⁵² Of note is that such scores indicate mild-moderate impairment on the measures employed, rather than “severe disability.”

Multiple regressions were performed to investigate predictors of outcome at 10 years post-injury for those variables that were statistically significant among the groups, as determined by the abovementioned ANOVAs. Correlations among independent variables were calculated, to identify multicollinearity. The Variance Inflation Factor (VIF) and the Condition Index (CI) were also calculated. VIF scores ranged between 1 and 2.64, indicating no problem with multicollinearity. Furthermore, only one variable had a CI score of>30, indicating possible collinearity problems. However, further exploration showed no variance proportion values of ≥0.50,⁵³ satisfying collinearity assumptions. Not unexpectedly, given the design of the study, age at injury and age at acute assessment correlated highly (r=0.99, p<0.001). Variables used in these analyses were: injury variable (acute GCS), social/environmental factor (acute SES, number of intervention services used over the 10 years), general intelligence during the acute stage (FSIQ), developmental factor (age at injury), and pre-injury family factor (FFQ-Intimacy Factor). Pre-injury adaptive ability (VABS – Total score) was also entered in Block 1 of the regression for the 10 year adaptive outcome measure, and pre-injury behaviour (PIC- Factor 1) was entered in Block 1 of the regression for the 10 year behavioral measure, with all other predictor variables entered in Block 2. By entering variables into Block 1 of the regression (those that may explain most of the variance), one is able to observe the contribution of other predictor variables (entered in Block 2) over and above that of the variables controlled for in Block 1.

Results

Sample characteristics

Comparison of the demographic and injury characteristics of participating and non-participating groups are provided in Table 1. Although no significant differences were found for children sustaining TBI, for control children a significant group difference for SES was identified, with those remaining in the study being of higher SES at time of recruitment (p=0.024). A significant difference was also evident for acute FSIQ, with participants in the study recording a higher score (p=0.003). The 10 year follow-up participants were also younger at first assessment than were non-participants (p<0.001).

Table 1.

Comparison of Participants and Non-Participants

	Seen at 10 years	Not seen at 10 years	p
TBI:	n=40	n=56
Age at injury (years), M (SD)	4.81 (1.92)	4.46 (2.07)	0.40
Male, n (%)	25 (62.50)	34 (60.71)	0.51
Intact families, n (%)	34 (85)	38 (69)	0.31
Family socioeconomic status, M (SD)	4.21 (0.93)	4.32 (1.11)	0.67
FSIQ-acute, M (SD)	99.29 (18.93)	97.98 (18.59)	0.76
VABS-acute, M (SD)	111.61 (19.02)	110.58 (18.44)	0.76
Acute GCS, M (SD)	9.65 (4.24)	9.11 (4.15)	0.54
Severity, n (%)			0.99
Mild	7 (17.5)	10 (17.9)
Moderate	20 (50.0)	28 (50.0)
Severe	13 (32.5)	18 (32.1)
Control group	n=19	n=16
Age at first assessment, M(SD)^**	4.06 (1.63)	6.23 (1.46)	< 0.001
Male, n (%)	12 (63)	6 (38)	0.12
Intact families, n (%)	17 (90)	13 (81)	0.45
Family socioeconomic status, M (SD)^*	3.39 (1.08)	4.30 (0.67)	0.024
FSIQ-acute, M (SD)^**	115.89 (17.46)	98.69 (12.91)	0.003
VABS-acute, M (SD)	113.83 (15.85)	113.0 (18.79)	0.902

p<0.01; ^* p<0.05.

GCS, Glasgow Coma Score; FSIQ, Full Scale Intellectual quotient; VABS, Vinleand Adaptive Behavior Scale; GCS, Glasgow Coma Score.

Forty children with TBI, (42%) of the original sample, took part in the 10 year review. Mean age at injury was 4.81 years (SD=1.92, age range of 1.17–8.00). Mean time since injury was 10.54 years (SD=1.57). Nineteen children (54%) of the original control group were seen at 10 year follow-up. As can be seen in Table 2, there were no significant differences across TBI and control groups with respect to age at acute or at 10 year assessment and for pre-injury adaptive abilities or family functioning. There was a significant difference between the moderate TBI group and the control group for SES, F(3,54)=3.52, p=0.02, and between the severe TBI group and the control group for FSIQ in the acute stage, F(3,55)=5.07, p=0.04. ANOVA also found a significant main effect for severity on FSIQ at 10 years F(3,58)=6.79, p=<0.01. Tukey's post-hoc test revealed that significant differences occurred between the moderate TBI group and the controls (p=0.04), and between the severe TBI group and controls (p<0.001).

Table 2.

Demographic Characteristics of the Sample

Demographics	Mild	Moderate	Severe	Control
n	7	20	13	19
Male, n (%)	3 (43)	13 (65)	9 (69)	12 (63)
Age at injury, M(SD)	4.68 (1.60)	5.07 (2.04)	4.47 (1.94)	n/a
Age at acute assessment (years), M (SD)	4.81 (1.61)	5.22 (2.02)	4.67 (1.93)	4.06 (1.63)
Age at 10 year follow-up (years), M (SD)	14.39 (1.28)	15.45 (3.36)	15.27 (2.96)	14.24 (2.29)
Range	12.9–16.9	10.3–21.2	11.3–20.4	11.5–19.5
Time lapse since injury, M (SD)	9.71 (1.64)	10.38–(1.83)	10.79 (1.43)	–
Range	7.9–12.3	7.4–13.3	8.9–13.3
Socioeconomic status acute, M (SD)^{* a}	4.17 (0.96)	4.31 (0.54)	4.08 (1.05)	3.40 (1.08)
VABS adaptive composite pre-injury, M (SD)	116.60 (26.43)	117.83 (13.47)	102.55 (18.67)	113.83 (15.85)
FSIQ acute, M (SD)^{** b}	98.00 (15.70)	103.10 (16.87)	90.77 (23.36)	115.89 (17.46)
FSIQ 10 years, M (SD)^** ^{a b}	98.86 (16.50)	102.40 (16.39)	91.85 (17.32)	116.26 (13.34)
FFQ intimacy pre-injury, M (SD)	60.57 (12.65)	65.85 (3.79)	63.46 (8.01)	66.84 (3.11)
FFQ intimacy 10 years, M (SD)	59.43 (12.65)	59.12 (12.65)	63.42 (5.98)	61.24 (7.22)
PIC-Factor 1, M (SD)	59.71 (7.48)	53.20 (9.04)	52.00 (6.66)	53.83 (17.97)
FBIS 10 years M (SD)^{* c}	4.29 (7.68)	9.78 (13.32)	21.54 (19.20)	n/a

Significant difference between the moderate and control groups.

Significant difference between the severe and control groups.

Significant difference between the mild and severe groups.

p<0.05; ^** p<0.01.

VABS, Vineland Adaptive Behavior Scale; FSIQ, Full Scale Intellectual Quotient; FFQ, Family Functioning Questionnaire; PIC, Personality Inventory for Children; FBIS, Family Burden Interview Scale.

As expected, there was a significant difference between injury severity TBI groups on injury characteristics such as the GCS, F(2,37)=17.43, p=<0.001); CT/MRI findings, χ² (2, n=40)=22.14, p<0.001; neurological signs, χ² (2, n=40)=16.45, p<0.001; surgical intervention, χ² (2, n=40)=10.27, p=< 0.01; duration of coma, χ² (2, n=40)=32.52, p<0.01. Children with severe injuries were more likely to demonstrate multiple areas of cerebral pathology, with a higher number presenting with neurological signs, surgical intervention. and injury via motor car accidents (Table 3).

Table 3.

Injury Characteristics of the 10 Year Sample

Injury characteristics	Mild	Moderate	Severe
n	7	20	13
Medical characteristics:
GCS on admission, M (SD)^a	13.57 (1.51)	10.85 (3.92)	5.69 (2.14)
Neurological signs, n (%)^a	0	9 (45)	12 (92)
Abnormal CT/MRI, n (%)^a	0	15 (75)	13 (100)
Surgical intervention, n (%)^a	0	6 (30)	9 (69)
Coma<1 day, n (%)^a	7 (100)	19 (95)	3 (23)
Cause of injury:
Passenger, n (%)	0	1 (5)	3 (23)
Pedestrian, n (%)	0	2 (10)	4 (31)
Falls, n (%)	7 (100)	14 (70)	3 (23)
Other, n (%)	0	3 (15)	3 (23)

Significant relationship between severity and injury characteristic.

GCS, Glasgow Coma Score.

Data were also collected for the children with TBI regarding the type of treatment/intervention they received across the 10 years post-injury including the percentage of the sample still accessing services at 10 years post-injury (Table 4). Whereas no participant received any specific attentional or cognitive training, the children who had sustained a severe TBI accessed more services over time and received the most educational assistance (69%), in comparison with those children who had sustained a moderate (28%) or mild (14%) TBI (Table 4).

Table 4.

Services Accessed by the Children from the Acute Stage to 10 Years Post-TBI

Services	Mild n=7	Moderate n=18	Severe n=13	p value
Neuropsychology, n (%)	0	1 (5.6)	10 (76.9)	<0.001
Psychology, n (%)	2 (28.6)	4 (22.2)	8 (61.5)	n/s
Physiotherapy, n (%)	1 (14.3)	2 (11.1)	10 (76.9)	<0.001
Speech pathology, n (%)	0	8 (44.4)	10 (76.9)	0.004
Educational, n (%)	1 (14.3)	5 (27.8)	9 (69.2)	0.021
Still receiving at 10 years, n (%)	1 (14.3)	5 (27.8)	7 (53.8)	n/s

Functional measures

Adaptive skills

For all ABAS Domains, Contrast 1 (severe TBI versus mild TBI and controls) was significant (Conceptual: t[52]=2.56, p=0.01; Social: t[52]=2.55, p=0.01; Practical: t[52]=1.81, p=0.04; Total Composite: t[52]=2.59; p=0.01) and Contrast 2 (moderate TBI versus mild TBI and controls) was non-significant (Conceptual: t[52]=1.54, p=0.07; Social: t [52]=1.61, p=0.06; Practical: t[52]=0.95, p=0.17; Total Composite: t[52]=1.43, p=0.08). Comparisons of impairment rates among the groups indicated that impairment levels were significantly higher at 10 years post-TBI for the severe group than for those with mild TBI and controls for ABAS Conceptual Domain, χ²(3, n=56)=9.42, p=0.02 and ABAS Social Domain, χ²(3, n=56)=9.42, p=0.02 (Table 5 for functional measures).

Table 5.

Analysis of Means (and Standard Deviations) of Functional Measures at 10 Years Post-TBI

	Controls M (SD)	Mild TBI M (SD)	Moderate TBI M (SD)	Severe TBI M (SD)	Effect size of planned contrasts^a
Adaptive skills
ABAS: Conceptual Domain^*	105.47 (12.23)	103.14 (14.98)	96.47 (19.94)	89.00 (14.50)	−1.19, −0.52
ABAS: Social Domain^*	102.37 (11.87)	104.71 (11.62)	95.95 (16.60)	89.45 (17.54)	−1.00, −0.51
ABAS: Practical Domain^*	98.53 (14.45)	98.57 (12.88)	92.84 (21.95)	85.82 (22.24)	−0.76, −0.32
ABAS: Total Composite^*	102.63 (13.24)	102.86 (15.26)	94.89 (20.00)	86.00 (19.38)	−1.08, −0.47
Behavior problems
BASC: Externalizing Problems Composite	46.44 (5.75)	59.00 (18.57)	49.35 (10.80)	48.18 (6.81)	−0.17, −0.05
BASC: Internalizing Problems Composite	46.84 (10.00)	56.43 (16.71)	48.76 (8.81)	55.17 (0.57)	0.48, −0.06
BASC: Behavioral Symptoms Index	46.42 (7.11)	55.43 (15.97)	49.59 (11.73)	55.85 (7.71)	0.71, 0.07
Educational skills
WRAT: Reading^*	113.95 (10.38)	98.14 (11.71)	99.90 (12.65)	100.17 (8.66)	−0.82, −0.77
WRAT: Spelling	114.74 (12.69)	96.00 (12.21)	102.00 (15.17)	99.83 (10.96)	−0.71, −0.51
WRAT: Arithmetic	101.47 (12.07)	88.14 (10.98)	88.35 (14.31)	89.25 (15.96)	−0.62, −0.70

Effect size=Cohen's d.

p<0.05.

Contrast 1=Severe vs. Mild/Control; Contrast 2=Moderate vs. Mild/Control; ABAS, Adaptive Behavior Assessment System; BASC, Behavior Assessment System for Children; WRAT, Wide Range Achievement Test.

Behavior problems

When investigating the BASC domains, no significant results were evident for Externalizing problems (Contrast 1, t[49]=1.10, p=0.15; Contrast 2, t[49]=0.76, p=0.23), Internalizing problems (Contrast 1, t[51]=−0.90, p=0.19; Contrast 2, t[51]=0.81, p=0.21), or for the Behavioral Index (Contrast 1, t[52]=−1.30, p=0.11; Contrast 2, t[52]=0.32, p=0.38). However, a number of ES were medium-large in this domain, suggestive of clinical significance. Impairment levels were significantly higher for the mild group than for the control group, and those with more severe injury for BASC Externalizing problems, χ²(3, n=53)=9.29, p=0.03.

Educational skills

Planned comparisons for reading ability revealed non- significant results for Contrast 1, t(54)=1.46, p=0.08, but significant results for Contrast 2, t(54)=1.76, p=0.04. With regard to spelling, neither Contrast 1, t(54)=1.15, p=0.13, nor Contrast 2, t(54)=0.81, p=0.21, were significant. Arithmetic followed a similar pattern with non-significant results for both Contrast 1, t(54)=1.12, p=0.13, and Contrast 2, t(54)=1.51, p=0.07. As indicated in Table 5, the ES for educational areas varied, ranging from medium to large. With regard to educational skills, χ² analysis did not reveal a significant difference in impairment rates between the groups for reading, χ²(3, n=58)=2.34, p=0.51, spelling, χ²(3, n=58)=3.11, p=0.38 or arithmetic, χ²(3, n=58)=5.25, p=0.16.

Predictors of outcome

For predictors of adaptive skills (See Table 6), pre-injury adaptive ability was entered in Block 1 (variable that may explain most of the variance) for all regression analyses. For the Conceptual Domain, Block 1 of the regression analysis revealed that pre-injury adaptive ability was a significant predictor of performance at 10 years post-injury, accounting for 32% of the variance. When other predictors were entered in Block 2, again pre-injury adaptive ability was a significant predictor (p=0.02) in addition to number of interventions since the accident (p=0.04), and these variables accounted for 58% of the variance. However, the R ² change was not significant, indicating that Block 2 of the regression did not significantly improve the amount of variance accounted for in Block 1. For the Social Domain, pre-injury adaptive ability was again a significant predictor, (accounting for 54% of the variance), with Block 2 including no other significant predictors. For the Practical Domain, only Block 1 was statistically significant, with pre-injury adaptive functioning accounting for 27% of the variance. With regard to the Total ABAS score, Block 1 of the regression analysis revealed that pre-injury adaptive function accounted for 36% of the variance, with Block 2 accounting for 54%, although no other significant predictors were found, nor was there a significant R ² change.

Table 6.

Predictors of Functional Outcome at 10 Years Post-TBI; B Hierarchical Regression Adjusting for Pre-Injury Adaptive

Predictor variables		ABAS conceptual	ABAS social	ABAS practical	ABAS total composite	WRAT reading
B1	Pre-injury Adaptive
	β, p-value	0.56, <0.01^*	0.73, <0.001^*	0.52, 0.01^*	0.60, <0.01^*	na, na
B2	Pre-injuryAdaptive, β, p-value	0.57, 0.02^*	0.63, 0.01^*	0.60, 0.03^*	0.56, 0.02^*	na, na
	GCS, β, p-value	−0.17, 0.48	0.09, 0.70	−0.07, 0.80	−0.06, 0.80	−0.11, 0.96
	SES, β, p-value	0.10, 0.56	−0.06, 0.71	−0.05, 0.80	−0.01, 0.95	−0.22, 0.89
	Number Intervention, β, p-value	−5.71, 0.04^*	−0.19, −0.45	−0.27, 0.36	−0.40, 0.15	0.07, 0.73
	Injury age, β, p-value	−0.02, 0.93	−0.08, 0.72	0.17, 0.49	0.08, 0.72	0.06, 0.75
	FSIQ, β, p-value	−0.26, 0.28	−0.04, 0.87	−0.24, 0.37	−0.14, 0.58	0.54, 0.01^*
	FF-Intimacy, β, p-value	0.17, 0.39	0.03, 0.88	0.28, 0.22	0.17, 0.42	−0.21, 0.20
	Model F	F(7,17)=3.33, p=0.02	F(7,17)=3.58, p<0.02	F(7,17)=2.05, p=1.11	F(7,17)=2.79, p=0.04	F(6,30)=1.87, p=0.12
	R ² Change^a	0.26	0.06	0.19	0.18	na

All R ² Change significance tests non-significant (p>0.05).

Significant predictors (p<0.05).

B1, Block 1; B2, Block 2; β, Unstandardized regression coefficients; GCS, Glasgow Coma Score; SES, Socioeconomic status; FSIQ, Full Scale Intellectual Quotient; ABAS, Adaptive Behavior Assessment System; WRAT, Wide Range Achievement Test, FF-Intimacy.Family Functioning-Intimacy.

With regard to reading outcome, although the overall model was not significant for Reading, R ²=0.27, Model F=F(6,30)=1.87, p=0.12, it was interesting to note that acute intellectual function was a significant predictor (p=0.007).

Discussion

Our findings partially supported the prediction that more severe TBI would be associated with greater functional deficits at 10 years post-injury. Adaptive functioning was influenced by injury severity, where those with more significant injury performed poorest. In contrast, behavioral outcomes were affected regardless of injury severity. With regard to educational performance, arithmetic was below the expected level for all severity groups 10 years post-TBI. Furthermore, 10 year adaptive outcomes were best predicted by pre-injury levels of child function, and educational ability by the acute stage post-injury intellectual function.

Adaptive outcomes

Study results are consistent with previous literature in which adaptive abilities in children were found to be influenced by injury severity.^25,31 There was a significant difference on all domains of adaptive function, with the severe TBI group performing poorest in comparison with those with lesser injury. In fact, the group mean for the severe group was close to one SD below the mean on the Practical and Total Composite Domains, unlike the mean scores of the other groups, which were close to the mean. Although no significant differences were found for Contrast 2, where the moderate group was compared with those in the mild and control groups, the pattern was similar, with those with more serious injury displaying the most difficulty in comparison with the other groups. Furthermore, ES ranged from medium to large, suggesting that more significant results may have been obtained with a larger sample size. Consistent with the group mean results, the number of children demonstrating impaired functional abilities was significantly greater for the severe TBI group for both Conceptual and Social Domains, also supporting past literature that suggests that such difficulties persist post-TBI.⁵⁴

Behavioral outcomes

Findings for the behavioral measure were not as anticipated. No significant differences were evident for externalizing or internalizing problems, or for behavioral symptoms among the severity groups, in contrast to previously reported studies.^5,7,8,28 These results do support literature stating that, unlike cognitive difficulties, which often show a dose–response relationship with injury severity,⁹ behavioral problems tend to be common post-TBI, regardless of injury severity.⁵⁵ In support of this latter view, the mildly injured group also showed a significantly higher percentage of survivors impaired for externalizing problems, and a trend toward a higher percentage impaired for internalizing problems and behavioral symptoms, again suggesting that behavioral difficulties may be evident regardless of injury severity. However, it is also worth noting that whereas we did exclude children with significant behavioral disorders, as some of the sample was recruited at a very young age, some of these problems may not have emerged at recruitment.

Educational outcomes

Our results were only partially consistent with previous research,^31,56,57 in which greater injury severity was linked to poorer educational performance in the areas of reading (decoding words), spelling, and arithmetic. In the current study, only Contrast 2 was significant for reading, indicating that the group with moderate TBI generally performed more poorly than those with milder or no injury; however, a larger sample size may have resulted in similar findings when comparing those with severe injuries with those with lesser injuries or no injuries, as suggested by the ES. It is of note that with regard to arithmetic, apart from the control group, all groups performed below the average range, supporting the claim that arithmetic may be most compromised post-injury, with this argument also supported by the higher percentage impaired for arithmetic compared with for reading and spelling.^58,59

Predictors

For adaptive outcome, pre-injury adaptive function was the strongest predictor, supporting past literature in highlighting the importance of pre-injury status,^{1,7,25,28,60
–62} with injury perhaps also exacerbating pre-injury status. Number of interventions since injury was also a significant predictor for the Conceptual Domain, indicating that factors external to the injury itself are influencing outcome, and also that the family's involvement and ability to access resources plays a role in recovery and longer-term outcome. Our results did not fully support past literature,^{1,25,56,57,63} in which injury severity and age at injury⁵⁴ were found to be significant predictors of educational outcome. It was interesting to note that acute intellectual functioning was a significant predictor of reading outcome, linking cognitive and educational skills.

Limitations and future directions

The findings for this study must be interpreted in the context of its methodological limitations. Regarding differences between participants and non-participants, whereas analysis revealed no significant differences for the TBI group, there were significant differences for the control group, in which participants were younger at first assessment, came from higher SES families, and had a higher IQ. To address this issue, SES and age at 10 years were employed in analyses. However, whereas the control group originally recruited was well matched to the injury group, the control group at 10 years may be characterized by some systematic bias.

Although few studies have systematically followed children with TBI for >5 years, such long-term studies are impacted by sample attrition; thus study results are limited because of the small sample size and unequal number of participants in each of the groups by 10 years post-recruitment, with a particularly small mild TBI group. Small sample size often does not allow for the use of more sophisticated statistical techniques. Further, there was a wide range with respect to time since injury. In general, TBI research suggests relatively stable function after early recovery and over the 2–3 years post-injury; therefore, a time since injury between 7 and 13 years should not be confounded by recovery processes. Furthermore, whereas the use of planned contrasts reduced the risk of type 1 error and allowed for directionality in the hypotheses, there was the possibility that significant differences that may have been evident using multivariate ANOVA and post-hoc analysis were missed. Additionally, recruitment age ranged from 1 to 8 years; therefore, providing accurate pre-injury functional data for the younger portion of the sample may have been difficult for parents.

Conclusion

The present study shows that functional difficulties do occur and persist to 10 years post-injury. In the adaptive skills area, results were as expected, where more serious TBI resulted in poorer adaptive ability and more impairment. In the behavioral area, results were not as hypothesized, but indicated that behavioural difficulties may occur regardless of injury severity. In the educational area, performance did not reveal a dose–response relationship for severity; however, poorer performance in arithmetic was evident by all groups with TBI. These results suggest that pre-injury functioning, intellectual ability, and family factors may all affect functional outcome in the long term, and they also revealed that parents of those with severe injury reported more burden. A more novel finding was the effect of number of interventions accessed in the prediction of outcome, emphasizing the importance of ensuring families are supported and guided in order to prevent or reduce the impact of childhood TBI.

Footnotes

Acknowledgments

This research was supported by the Australian National Health and Medical Research Council, the Royal Children's Hospital Research Foundation, and the Victorian Government's Operational Infrastructure Support Program.

Author Disclosure Statement

No competing financial interests exist.

References

Catroppa

, Anderson

V.A.

, Morse

S.A.

, Haritou

, Rosenfeld

J.V.

2008. Outcome and predictors of functional recovery 5 years following pediatric traumatic brain injury (TBI) J. Pediatr. Psychol., 33:707–718.

Groom

K.N.

, Shaw

T.G.

, O'Connor

M.E.

, Howard

N.I.

, Pickens

1998. Neurobehavioral symptoms and family functioning in traumatically brain-injured adults. Arch. Clin. Neuropsychol., 13:695–711.

Klonoff

1971. Head injuries in children: predisposing factors, accident conditions, accident proneness and sequelae. Am. J. Public Health, 61:2405–2417.

Catroppa

, Anderson

2007. Recovery in memory function, and its relationship to academic success, at 24 months following pediatric TBI. Child Neuropsychol., 13:240–261.

Chapman

L.A.

, Wade

S.L.

, Walz

N.C.

, Taylor

H.G.

, Stancin

, Yeates

K.O.

2010. Clinically significant behavior problems during the initial 18 months following early childhood traumatic brain injury. Rehab. Psychol., 55:48–57.

Thomsen

I.V.

1984. Late outcome of very severe blunt head trauma: a 10–15 year second follow-up. J. Neurol. Neurosurg. Psychiatry, 47:260–268.

Yeates

K.O.

, Swift

, Taylor

H.G.

, Wade

S.L.

, Drotar

, Stancin

, Minich

2004. Short- and long-term social outcomes following pediatric traumatic brain injury. J. Int. Neuropsychol. Soc., 10:412–426.

Yeates

K.O.

, Taylor

H.G.

, Chertkoff Walz

, Stancin

, Wade

S.L.

2010. The family environment as a moderator of psychosocial outcomes following traumatic brain injury in young children. Neuropsychology, 24:345–356.

Anderson

, Brown

, Newitt

, Hoile

2009. Educational, vocational, psychosocial, and quality-of-life outcomes for adult survivors of childhood traumatic brain injury. J. Head Trauma Rehabil., 24:303–312.

10.

Max

, Koele

S.L.

, Smith

W.J.

, Sato

, Lindgren

, Robin

, Arndt

1998. Psychiatric disorders in children and adolescents after severe traumatic brain injury: a controlled study. J. Am. Acad. Child Adolesc. Psychiatry, 37:832–840.

11.

Max

, Robin

D.A.

, Lindgren

S.D.

, Smith

W.L.

, Sato

, Mattheis

P.J.

, Stierwalt

J.A.G.

, Castillo

C.S.

1997. Traumatic brain injury in children and adolescents: psychiatric disorders at two years. J. Am. Acad. Child Adolesc. Psychiatry, 39:1278–1285.

12.

Rivara

, Jaffe

, Polissar

, Fay

, Martin

, Shurtleff

, Liao

1994. Family functioning and children's academic performance in the year following traumatic brain injury. Arch. Phys. Med. Rehabil., 75:369–379.

13.

Taylor

H.G.

, Yeates

K.O.

, Wade

S.L.

, Drotar

, Stancin

, Minich

2002. A prospective study of short- and long-term outcomes after traumatic brain injury in children: behavior and achievement. Neuropsychology, 16:15–27.

14.

Andrews

T.K.

, Rose

F.D.

, Johnson

D.A.

1998. Social and behavioural effects of traumatic brain injury in children. Brain Inj., 12:133–138.

15.

Goldstrohm

S.L.

, Arffa

2005. Preschool children with mild to moderate traumatic brain injury: an exploration of immediate and post-acute morbidity. Arch. Clin. Neuropsychol., 20:675–695.

16.

Max

, Roberts

M.A.

, Koele

S.L.

, Lindgren

S.D.

, Robin

D.A.

, Arndt

, Smith

W.L.J.

, Sato

1999. Cognitive outcome in children and adolescents following severe traumatic brain injury: influence of psychosocial, psychiatric and injury-related variables. J. Int. Neuropsychol. Soc., 5:58–68.

17.

Max

, Lansing

A.E.

, Koele

S.L.

, Castillo

C.S.

, Bokura

, Schachar

2004. Attention deficit hyperactivity disorder in children and adolescents following traumatic brain injury. Dev. Neuropsychol., 25:159–177.

18.

Yeates

K.O.

, Taylor

H.G.

, Wade

S.L.

, Drotar

, Stancin

, Minich

2002. A prospective study of short- and long-term neuropsychological outcomes after traumatic brain injury in children. Neuropsychology, 16:514–523.

19.

Dooley

J.J.

, Anderson

, Hemphill

S.A.

, Ohan

2008. Aggression after pediatric traumatic brain injury: a theoretical approach. Brain Inj., 22:836–846.

20.

Fay

, Yeates

K.O.

, Wade

S.L.

, Drotar

, Stancin

, Taylor

G.H.

2009. Predicting longitudinal patterns of functional deficits in children with traumatic brain injury. Neuropsychol. Rev., 23:271–282.

21.

Anderson

, Moore

1995. Age at injury as a predictor of outcome following pediatric head injury: a longitudinal perspective. Child Neuropsychol., 1:187–202.

22.

Barnes

M.A.

, Dennis

, Wilkinson

1999. Reading after closed head injury in childhood: effects on accuracy, fluency and comprehension. Dev. Neuropsychol., 15:1–24.

23.

Gronwall

, Wrightson

, McGinn

1997. Effect of mild head injury during the preschool years. J. Int. Neuropsychol. Soc., 3:592–597.

24.

Wrightson

, McGinn

, Gronwall

1995. Mild head injury in preschool children: evidence that it can be associated with a persisting cognitive deficit. J. Neurol. Neurosurg. Psychiatry, 59:376–380.

25.

Anderson

, Catroppa

, Haritou

, Morse

, Rosenfeld

2005. Identifying factors contributing to child and family outcome at 30 months following traumatic brain injury in children. J. Neurol. Neurosurg. Psychiatry, 76:401–408.

26.

Anderson

, Catroppa

, Dudgeon

, Morse

, Haritou

, Rosenfeld

2006. Understanding predictors of functional recovery and outcome thirty-months following early childhood head injury. Neuropsychology, 20:42–57.

27.

Wells

, Minnes

, Phillips

2009. Predicting social and functional outcomes for individuals sustaining paediatric traumatic brain injury. Dev. Neurorehabil., 12:12–23.

28.

Gerring

J.P.

, Grados

M.A.

, Slomine

, Christensen

J.R.

, Salorio

C.F.

, Cole

W.R.

, Vasa

R.A.

2009. Disruptive behaviour disorders and disruptive symptoms after severe paediatric traumatic brain injury. Brain Inj., 23:944–955.

29.

Alves

1992. Natural history of post-concussive signs and symptoms. Phys. Med. Rehabil. State Arts Rev., 6:21–32.

30.

Donders

, Nesbit–Greene

2004. Predictors of neuropsychological test performance after pediatric traumatic brain injury. Assessment, 11:275–284.

31.

Ewing–Cobbs

, Prasad

M.R.

, Kramer

, Cox

C.S.

, Baumgartner

, Fletcher

, Mendez

, Barnes

, Zhang

, Swank

2006. Late intellectual and academic outcomes following traumatic brain injury sustained during early childhood. J. Neurosurg., 105:287–296.

32.

Rivara

J.B.

, Jaffe

K.M.

, Fay

G.C.

, Polissar

N.L.

, Martin

K.M.

, Shurtleff

H.A.

, Liao

1993. Family functioning and injury severity as predictors of child functioning one year following traumatic brain injury. Arch. Phys. Med. Rehabil., 74:1047–1055.

33.

Stuss

, Anderson

2004. The frontal lobes and theory of mind: developmental concepts from adult focal lesion research. Brain and Cogn., 55:69–83.

34.

Taylor

H.G.

, Drotar

, Wade

S.L.

, Yeates

K.O.

, Stancin

, Klein

1995. Recovery from traumatic brain injury in children: the importance of the family. Traumatic Head Injury in Children. Broman

S.H.

, Michel

M.E.

Oxford University Press: New York, 188–218.

35.

Wade

S.L.

, Taylor

H.G.

, Yeates

K.O.

, Drotar

, Stancin

, Minich

N.M.

, Schluchter

2006. Long-term parental adaptation and family adaptation following pediatric brain injury J. Pediatr Psychol., 31:1072–1083.

36.

Anderson

, Morse

, Klug

, Catroppa

, Haritou

, Rosenfeld

J.V.

, Pentland

1997. Predicting recovery from head injury in young children: a prospective analysis. J. Int. Neuropsychol. Soc, 3:568–580.

37.

Anderson

, Catroppa

, Morse

, Haritou

, Rosenfeld

2000. Recovery of intellectual ability following traumatic brain injury in childhood: impact of injury severity and age at injury. Pediatr. Neurosurg, 32:282–290.

38.

Teasdale

, Jennett

1974. Assessment of coma and impaired consciousness. Lancet, 2:81–84.

39.

Yeates

K.O.

, Armstrong

, Janusz

J.A.

, Taylor

H.G.

, Wade

S.L.

, Stancin

, Drotar

2005. Long-term attention problems in children with traumatic brain injury. J. Am. Acad. Child Adolesc. Psychiatry, 44:574–584.

40.

Daniel

1983. Power, Privilege and Prestige: Occupations in Australia. Longman–Cheshire: Melbourne.

41.

Sparrow

, Balla

D.A.

, Cicchetti

D.V.

1984. Vineland Adaptive Behavior Scales: Interview Edition. Survey Form Manual. American Guidance Services: Circle Pines, MN.

42.

Harrison

P.L.

, Oakland

2003. Adaptive Behavior Assessment System. San Antonio, TX: Harcourt Assessment.

43.

Lachar

1992. Personality Inventory for Children (PIC) Western Psychological Services: Los Angeles.

44.

Reynolds

C.R.

, Kamphaus

R.W.

1992. Behavior Assessment System for Children. Guilford Press: New York.

45.

Noller

1988. ICPS Family Functioning Scales. University of Queensland.

46.

Bayley

1969. Bayley Scales of Infant Development: Birth to Two Years. The Psychological Corporation: San Antonio.

47.

Wechsler

1987. Manual for the Preschool and Primary Intelligence Scale–Revised. The Psychological Corporation: New York.

48.

Wechsler

1991. Wechsler Intelligence Scale for Children – Third Edition. The Psychological Corporation: San Antonio.

49.

Wechsler

1997. Wechsler Adult Intelligence Scale – Third Edition. The Psychological Corporation: San Antonio.

50.

Wilkinson

G.S.

1993. Wide Range Achievement Test Administration Manual. Wide Range Inc: Wilmington, DE.

51.

Cohen

. 1988. Statistical Power Analysis for the Behavioral Sciences. Academic Press: San Diego.

52.

Tabachnick

B.G.

, Fidell

L.S.

2001. Using Multivariate Statistics. Allyn and Bacon: Needham Heights, MA.

53.

Rawlings

J.O

, Pantula

S.G

, Dickey

D.A

. 1998. Applied Regression Analysis:A Research Tool, 2nd. Springer: New York.

54.

Muscara

, Catroppa

, Eren

, Anderson

2009. The impact of injury severity on long-term social outcome following paediatric traumatic brain injury. Neuropsychol. Rehabil., 19:541–561.

55.

Woods

, Catroppa

, Giallo

, Matthews

, Anderson

2012. Feasibility and consumer satisfaction ratings following an intervention for families who have a child with acquired brain injury. Neurorehabil., 30:189–198.

56.

Catroppa

, Anderson

V.A.

, Muscara

, Morse

S.A.

, Haritou

, Rosenfeld

J.V.

, Heinrich

L.M.

2009. Educational skills: long-term outcome and predictors following paediatric traumatic brain injury. Neuropsychol. Rehabil., 19:716–732.

57.

Ewing–Cobbs

, Barnes

, Fletcher

J.M.

, Levin

H.S.

, Swank

P.R.

, Song

2004. Modeling of longitudinal academic achievement scores after pediatric traumatic brain injury. Dev. Neuropsychol., 25:107–133.

58.

Berger–Gross

, Shackelford

1985. Closed-head injury in children: Neuropsychological and scholastic outcomes. Percept. Mot. Skills, 61:254–254.

59.

Slater

E.J.

, Kohr

M.A.

1989. Academic and intellectual functioning of adolescents with closed head injury. J. Adolesc. Res., 4:371–384.

60.

Cattelani

, Lombardi

, Brianti

, Mazzucchi

1998. Traumatic brain injury in childhood: intellectual, behavioural and social outcome into adulthood. Brain Inj, 12:283–296.

61.

Max

, Robin

D. A.

, Lindgren

S. D.

, Smith

W. L.

, Sato

, Mattheis

P. J.

, Stierwalt

J.A.

, Castillo

C.S.

1998. Traumatic brain injury in children and adolescents: Psychiatric disorders at one year. J. Neuropsychiatr., 10:290–297.

62.

Yeates

K.O.

, Taylor

H.G.

, Drotar

, Wade

S.L.

, Stancin

, Klein

1997. Pre-injury environment as a determinant of recovery from traumatic brain injuries in school-aged children. J. Int Neuropsychol. Soc., 3:617–630.

63.

Fletcher

J.M.

, Ewing–Cobbs

, Miner

M.E.

, Levin

H.S.

1990. Behavioral changes after closed head injury in children. J. Consult. Clin. Psychol., 58:93–98.